Introduction

The multiple alternate fractures of the spine are defined as those that present injuries at several levels separated by at least one healthy vertebral body. They were initially described in 1976 [1]. In the era before the computed tomography scan (CT scan), less than 8% [2] prevalence of injuries were reported. More recent publications reported a prevalence of up to 23% after a traumatic injury [3,4,5]. The increase in the prevalence can result from an increase in traffic accidents or a better diagnosis [5].

The main production mechanisms are the following: falls from height, traffic accidents, and direct traumas. It is thought that this kind of injuries presents complex production mechanisms, and they are related to more serious injuries [3, 5]. Up to 45.5% of association with brain injury [6, 7] were reported.

For the second injury of the spine, a delay in the diagnosis occurs in 23–28% of the cases; this can be related to an increase in pain, deformity, and distant neurological injuries, up to 9% [5].

Given the high prevalence of multiple fractures reported in the literature, we believe it is necessary to know the rates in our population of patients. We have not found reports in our country. Therefore, the purpose of this study is looking for the prevalence of multiple alternate spine fractures in patients that have been admitted in the emergency rooms of our five hospitals in the last 10 years and to set their relationship with other characteristics of interest.

Materials and Method

An observational descriptive retrospective study was conducted from April 1, 2007, to April 1, 2017, using medical records and scans of patients with spine fractures.

Information of different medical care centers of the City of Buenos Aires was included, such as Hospital General de Agudos C. Durand, Hospital Militar Central Buenos Aires Dr. C. Argerich, Hospital de Clínicas de Buenos Aires Gral. José de San Martín (UBA), Sanatorio Municipal Dr. Julio Méndez, and Sanatorio V. Franchin.

Due to the retrospective nature of the study, the fact that data was collected from the routine records of patients of each of the participating centers and that the filiation information of patients included was preserved, the informed consent of patients was not necessary. This study was evaluated and accepted by the ethics committee of the “Sociedad Argentina de Patología de la Columna Vertebral” (Argentinian Society of Spine Pathology), according to the current national regulations.

Variables

An alternate spine fracture was defined as the one that present a lesion seated in several levels of the spine, separated by at least one healthy vertebral body. A contiguous spine fracture was considered one that presents one or more spine fractures not separated by a healthy vertebral body [1].

In addition, the following information was considered: sex, age, production mechanism, neurological injury, resolution (surgical or not), injured region, related injuries, and requested testing.

Inclusion Criteria

Acute fractures of the spine: those affecting the patient attending consultation, with at least one traumatic history of no more than 1 week of evolution, related to pain, deformity and/or neurological compromise, identifying a structural or signal alteration with image methods.

Exclusion criteria

  • Subacute or chronic fractures with more than 1 week of evolution

  • Pathological fractures

  • Incomplete medical record, defined as the lack of two or more items of study interest

  • Presence of congenital anomalies

  • Loss of consultation

  • Related pathologies (diffuse, tumoral hyperostosis, ankylosing diseases)

Classifications

The classifications of the AO were taken for cervical subaxial [8] and thoracolumbar [9] injuries, without forgetting the Argenson and Mc Lain stability criteria [10,11,12,13,14]. Anderson classification was applied [15, 16] to odontoid fractures, Effendi Laurin classification [17] to C2 pedicle ones, and Geweiler classification to C1 [18, 19]. ASIA [20, 21] and Frankel [22] were applied to neurological assessments.

Statistical Analysis

Two groups were organized for data analysis: the “Alternate” group (multiple alternate spine fractures) and the “Contiguous” group (simple and multiple contiguous spine fractures).

The continuous variables assuming a normal distribution were reported as mean and standard deviation (SD). Otherwise, the median and interquartile ranges (IQR) were applied. The categorical variables were reported as presentation and percentage number. To determine the distribution of the sample, the Shapiro-Wilk or Kolmogorov Smirnov test was used, as appropriate.

To compare the continuous variables, the “Student’s t-test or the Mann-Whitney U Test” was used, as appropriate. To compare the categorical variables, the Chi2-test or Fisher’s exact test was used, as appropriate.

A p value < 0.05 was deemed significant.

The IBM SPSS Macintosh software, version 24.0 (IBM Corp., Armonk, NY, USA), was used for data analysis.

Results

A total of 514 medical records of the participating medical care centers were reviewed, within a period from April 2007 to April 2017. One hundred and twenty patients met the screening criteria and were included in the final analysis.

Of the total of spine fractures, 56 (46.7%) were simple, 35 (29.2%) were multiple contiguous, and 29 (24.1%) were multiple alternate (Table 1).

Table 1 Analysis of the general characteristics of the sample
Full size table

Fifty-four (45%) patients were women, and the average age of the sample was 40 years old (IQR 26.2–57.7). The rest of patients’ characteristics are described in Table 1.

In the Alternate group, the most affected area was the cervical one in 16 (55.1%) patients. In the Contiguous group, the most affected area was the lumbar one in 55 (49.4%) patients. Images 1, 2 depict the kind of fractures according to AO classification for the cervical subaxial (Image 1) and thoracolumbar (Image 2) subgroups.

Image 1
figure 1

Frequency of distribution of cervical subaxial injuries according to AO classification, in absolute numbers of patients

Full size image
Image 2
figure 2

Frequency of distribution of thoracolumbar injuries according to AO classification, in absolute numbers of patients

Full size image

Nine (31%) patients of the Alternate group presented an injury related to a labor accident, compared to 9 (9.9%) patients of the Contiguous group. The differences were statistically significant (p = 0.02).

Neurological injury: ten patients (34.5%) of the Alternate group presented a type A injury (Frenkel [22]) in comparison to nine (10.2%) patients of the Contiguous group, being statistically significant (p = 0.003); this was not the case in comparisons between groups for the type of injuries of B to E (Table 1).

Injury mechanism: no statistically significant differences were observed (p = 0.31).

Associated extra-spinal injury: 14 (48.3%) patients of the Alternate group presented at least one, in comparison to 23 (25.3%) patients of the Contiguous group. This is statistically significant (p = 0.02) (Table 1).

Ninety-two (76.7%) patients received surgical treatment, 21 (72.4%) patients from the Alternate group and 71 (78%) patients from the Contiguous group (p = 0.53). The carried-out therapy is described in Table 2. The average time before the surgery was 12 (IQR 3–21.2) days for the Contiguous group and 4 (IQR 2–11) days for the Alternate group (p = 0.058) (Table 1).

Table 2 Type of treatment applied
Full size table

From the total number of patients, 4 (3.3%) patients died, 2 (2.2%) patients from the Contiguous group (one with a high cervical injury and respiratory arrest and the other due to complications with a heart trauma), and 2 (6.9%) patients from the Alternate group (p = 0.25) (one due to Candida sepsis and the other due to complications of thoracic trauma).

Discussion

Our work allowed us to register a prevalence of multiple alternate fractures of 24.1%, slightly exceeding the international literature (17 or 19%) [4, 5, 23, 24], with similar production mechanisms, mainly in men, of an average age of 40 years old, slightly below of what was reported [3, 5, 23].

The anatomical location patterns of these fractures are similar to what was reported in the world literature: cervico-thoracic predominance (24.1%) followed by thoracolumbar (17.2%).

A case report, of L.Y. Dai y L.S. Jia [3], about 875 patients shows a prevalence of 5.4% (n = 47) of multiple alternate fractures, being similar to our sample in distribution per sex and average age (44.5 years old) and injury mechanism; while these have the thoracolumbar association as the most frequent, ours is the cervico-thoracic.

Miller et al. [4] report in their retrospective study 19% alternate injuries in patients with a cervical fracture, being the cervico-thoracic association as the most frequent one, and 57% patients presented extra-spinal injuries (48.3% of this study), with a higher prevalence in those patients with C7 injuries. They suggest that, in the presence of cervical injuries, the complete evaluation of the spine is mandatory to decrease the delay in the diagnosis of alternate fractures, especially in unconscious patients.

An extensive review of Nelson et al. [5] on 83,338 patients with fractures of the spine, due to direct trauma, reported 19% alternate injuries, 9% of which were cervico-thoracic, 4% cervico-lumbar, and 10% thoracolumbar. The male sex predominates (67%), with injuries related to serious trunk ones, mainly of the thorax (22%) with greater morbimortality, like this study where we found a relation of 24.1% thoracic injuries (without statistical significance). They were associated to high-energy traumas. According to the author, more multiple alternate fractures with high-energy mechanisms rather than low-energy ones were presented (60% likelihood increase). They reported a statistically significant difference in relation to the mortality among the high- and low-energy traumas (6.7% vs. 5.3%). They establish the presence of a diagnosed cervical fracture and the high-energy mechanism as risk factors of alternate injuries; especially in those related to traffic accidents, with the cervico-thoracic pattern as the most frequent (Table 3).

Table 3 Data compared to international bibliography
Full size table

A retrospective study (Wang et al. [23]) selected 213 patients with multiple alternate injuries of the spine. It highlighted a male predominance (57%) of patients under 60 years old (young patients) (67%), with a greater number of falls from height and traffic accidents, whereas among older patients, falls from low height and osteoporotic fractures predominated. Pathological fractures have been excluded from our retrospective study, among which the osteoporotic can be found. The thoracolumbar association was the most frequent, being the cervico-thoracic more frequent among young patients. The percentage of neurological injuries was 45.5%, with a significant difference in frequency between young people and elderly people (57.3 vs. 21.4%). Associated injuries 38.5%, the thoracic and skull injuries were the most frequent. The 66% of them were surgically solved. They propose the need of completely evaluating the spine, initially with front and profile X-rays, and, in case of injuries, complete the study with a CT scan (Tables 1 y 3).

A recent publication, of Kanna and col. [24], establishes that the diagnose of these fractures varies according to the complementary study applied, which is lower in simple X-rays and higher when CT scans and magnetic resonances (NMR) are involved. It agrees with the previous reports in that the delay in the diagnosis can lead to sequelae and an increase in neurological injuries. It states that the NMR of the complete spine helps to reduce the error in the diagnosis of secondary injuries, but they also highlight its limitation for polytraumatized patients, both due to the access to the NMR scanner and as they might be unstable to perform it. It has an important relation with this study, in which all were submitted to a CT scan and 96.6% to a NMR scan, which has increased sensitivity, especially in patients with intra-spongy fractures, in which there is little or no deformity shown by X-ray and tomography, but an important edema is seen in the NMR (Table 1).

Surgical therapy is present from the classic works of Robinson [25], going through the works of Cloward [26] and Sicard [27], up to the modern works of Kanna [24], Mourelo Fariña [28], and even the work of Steverlynk [29] and col. (with the use of titanium material in thoracolumbar fractures), and it is imposed on patients with instability. This study shows us that we have given mixed therapeutic responses to the same patient, with 72.4% of surgical resolutions, but bloodless treatment for type A spinal injuries (AO classification) (Table 2).

We can observe in the results that the presence of neurological injuries and associated extra-spinal injuries was higher in the Alternate group, being statistically significant. We suggest that in the case of a patient with spine fracture associated with a neurological injury or extra-spinal injuries, a second injury of the spine should be suspected (Table 1).

The association we found among the multiple alternate injuries and labor accidents (the majority being men in productive age) should alert us looking for a second injury when we are in presence of this kind of patient (Table 1).

This work has some limitations. In the first place, we can list the possible biases due to the retrospective nature of the study. In the second place, the weaknesses imposed from the use of different centers with multiple evaluators and treating teams. And finally, we should explain that not all the centers had an NMR scanner, so this could have underestimated the reported prevalence. Because of the chosen methodological design, the results should be considered cautiously, due to other potential factors or variables that may influence them.

As a strength, we can say that this is the first report of this kind of fractures in our country. We consider of importance the knowledge of this pathology for an adequate management of the patient.

A prospective study considering the data reported in this publication could be useful to systematize a unified therapeutic protocol.

Conclusion

One in four patients admitted in our centers with a spine fracture had at least other vertebrae fractured at a distance from the first one.

Those patients presented statistically significant associations with neurological injuries, extra-spinal injuries, and labor accidents. We suggest reevaluating the patients presenting these characteristics to reduce the error rate and a delay in the diagnosis.